PPE Shortages and the Failure to Increase Prices

Ryan Peterson, Flexport Founder and CEO, has an excellent piece on Why There Aren’t Enough Masks, and How to Get More. One part of the problem is a lack of working capital brought about in part by a fear of raising prices:

Typically, buyers of PPE, whether hospitals or medical distributors, expect to place purchase orders and only pay for products upon delivery, or even later.

But when demand surges by 20x, vendors simply don’t have the money required to scale production. Factories need money to add production lines, buy raw materials, and hire workers. They need down payments so they can move.

Buyers prefer to pay upon receipt of goods for two reasons. The first is to ensure quality: They can refuse payment if the goods they receive don’t meet their standards. The second reason is they prefer to keep cash on their balance sheets, rather than paying vendors in advance.

In ordinary times, sellers will accept this. But with the entire world desperate to buy PPE, manufacturers know they can ask for a down payment and get it. Other more aggressive entities are paying down payments, so if US buyers won’t, they don’t get the supply.

American medical distributors, governments, and even hospital chains, by contrast, have been less willing, or less able, to adapt to the new reality of paying vendors upfront, at higher prices than they’d contracted.

At the same time, US distributors can’t pass higher prices through to hospitals in the midst of the crisis, for fear of being accused of profiteering. Foreign governments and healthcare systems have been less encumbered by this, showing a willingness to pay more and pay faster to get first in line.

There was a recent debate on twitter about so-called price-gouging. It was said that the argument for raising prices is weak when the elasticity of supply is low. That’s not necessarily true. First, in an emergency even a small increase in quantity can be very valuable so high prices can have high utility payoffs. Second, vendors face credit market frictions and capital constraints. Borrowing in an emergency is often not possible–this means that asset balances matter and transferring wealth from buyers to firms can ease financial constraints. Put another way, it’s the short run increase in price which allows long run elasticities to increase. Elasticity is endogenous to pricing.

Hat tip: Paul Graham.


.... Chinese sellers now have to worry about the CCP coming down on them like a ton of lead due to the sort of inferior goods revealed by buyer testing. Such as substandard surgical gowns sold to the UK, for a more recent example.

Which means it sounds like American buyers are being quite smart in terms of this quote 'American medical distributors, governments, and even hospital chains, by contrast, have been less willing, or less able, to adapt to the new reality of paying vendors upfront, at higher prices than they’d contracted.'

Elasticity may be endogenous to pricing, but testing the quality of goods goes back to the roots of caveat emptor. Which wiki does a nice job explaining, if perhaps not to the satisfaction of economists with high GRE scores.

Caveat emptor (/ˈɛmptɔːr/; from caveat, "may he beware", a subjunctive form of cavēre, "to beware" + ēmptor, "buyer") is Latin for "Let the buyer beware".[1] It has become a proverb in English. Generally, caveat emptor is the contract law principle that controls the sale of real property after the date of closing, but may also apply to sales of other goods. The phrase caveat emptor and its use as a disclaimer of warranties arise from the fact that buyers typically have less information than the seller about the good or service they are purchasing. This quality of the situation is known as 'information asymmetry'. Defects in the good or service may be hidden from the buyer, and only known to the seller.

At present, the alternative to poor quality PPE is no PPE.

Unless one is really, really desperate right now, but certainly any nation watching what was happening back in February is well equipped to deal with the then quite foreseeable problems of coronavirus.

Back in February we were told that our PPE burden was not going to be substantially increased. The WHO's 2019 Novel Coronavirus (2019-nCoV) Strategic Preparedness and Response Plan has no recommendation for increased procurement of PPE, it anticipates that only undersourced health systems will have difficulties with securing infection control materials. Their official estimates for the increased cash needed for all increased medical expenditures was $675,680,427 (this includes masks, N-95s, ventilators, vaccine research, and everything else).

We were watching. We sought specific guidance from the experts. Their official conclusion was that we did not additional protective gear and that, at most, we would need to drop a bunch of cash to resource developing nations.

For all the retrospective prognostication I was in the bloody meeting where we were told that we would not need to stockpile.

Is this Dominic Cummings?

Undeniably believe that that you stated. Your favourite justification seemed
to be at the internet the easiest thing to remember of.
I say to you, I certainly get annoyed while other folks think about worries that they
plainly don't understand about. You managed to hit the nail upon the top and defined out the entire
thing without having side-effects , people could take a signal.
Will likely be again to get more. Thank you

Quite foreseeable, starting after the end of Fasching/Carnival. But then, America is different, and as Italy started to suffer, it was clear that Italy was not America. And the EU wasn't America, etc.

The basic point was that no one was interested in a problem that they thought containable, which by February, was clearly not the case unless one was willing to use Chinese methods. And even then, it was obvious that the Chinese were using massive amounts of resources to deal with their outbreak.

You can blame anyone you wish, there being plenty of blame to spread around, but this was also the reality back in March, where what was happening in Italy was completely impossible to ignore, unless one worked at it. From an article published back on March 21, when truly there was no excuse for only one person raising the alarm in mid-March or so - "His hospital had not prepared for this volume before the virus first appeared. One physician had tried to raise alarms, asking about negative pressure rooms and ventilators. Most staff concluded that he was overreacting. “They thought the media was overhyping it,” the respiratory therapist told me. “In retrospect, he was right to be concerned.”"

We do not trust our guts. Our system has a dedicated protocol for epidemics. Part of that is estimating PPE burden. One of the key questions for us is if we will need to go to universal precautions. That is expensive and ideally we need to place our PPE order two months before we need it.

ID became concerned back in January, they did their due diligence. They activated our early epidemic contingency plan. We talked directly to the CDC, ID read everything that the WHO was putting out. We plugged numbers into our formulas. We made a modest increase in our PPE purchases.

Now we are not hit too hard, but have gone to universal precautions nonetheless. We are only keeping afloat on PPE by taking extraordinary measures. Using the best on the ground evidence at the time, and remembering that the wrong call risks our lives, we got it wrong. The sources we trusted failed us.

Given what was known when made our decisions back then, it was not foreseeable if you accepted WHO recommendations at face value (which it appears the CDC did and we trusted the CDC). Even if you ignored the WHO, it is not easily foreseeable for the vast majority of specialist physicians to project this outcome. Thanks to the lying by the CCP we, at best, had to make immediate decisions with scant data or they became untimely. All of which was, of course, made worse by Chinese double dealing (e.g. they bought out one of our suppliers via a front company when they still were telling us that the vast majority of spread is from symptomatic patients). End of Carnival was what, the 26th? Yeah, maybe it was foreseeable then, but the vast bulk of PPE was already gone and the process takes time regardless of how much you hurry.

'We do not trust our guts. Our system has a dedicated protocol for epidemics.'

Except it clearly really didn't, at least from what you write. You blame the WHO and CCP extensively, for example, but neither is part of the American government, and neither run any American hospital chains.

And you completely ignore the political context, which was obliquely noted by that article - all during February and into March, the American government was often denying that there was a pandemic to be concerned about at all. If denial is part of that dedicated American epidemic protocol, well, no contradiction from me. And if denial isn't a part, then something went seriously wrong by not following a plan intended to preserve the health of Americans.

Our hospital system has an explicit dedicated protocol. As a private entity we take our responsibilities seriously and thus do not trust our guts. We activated that protocol back in January after the head of ID got worried. It has a low threshold to trip, precisely because we do not want to get burned.

The head of ID is a professional and knows his things, but he is based in the states and not in China. China was the only place with deaths so we went to read data that China has promised, and if I recall correctly is obligated by treaty, to provide fully and accurately. They gave data to the WHO. The WHO put out an official bulletin. In it you can read many things. It denigrates travel bans as ineffective, even though my contacts with MSF who actually staffed the Ebola epidemic beg to differ (and the public health organizations of every country that actually beat the curve up till now). They had their estimates of the costs needed to be spent, they are already an order of magnitude too low.

We made a good faith effort. We looked at the places that international treaty demands are supposed to have actionable intelligence. Our tax dollars cover a huge portion of their budget.

And we got it wrong.

This was not an easy call that was missed. It was one where if you followed best practices, you missed it because China lied.

> For all the retrospective prognostication I was in the bloody meeting where we were told that we would not need to stockpile.

We see stories on the news of hospitals doing massive pandemic and earthquake training. Not a single doctor at the hospital has ever said "Hey if this is real, I'm going to need 30 sets of PPE per day for a pandemic, times 500 doctors and nurses. That's 15K per day for our hospital, times 30 days, that's half a million sets of PPE at our hospital in the first month of a pandemic. Times 5 hospitals per big city, and 100 big cites, that's 250M sets of PPE the first month of a pandemic. Do we have this?"

How in the hell does a head doctor at a hospital not have this most basic analytical capacity? In other words, if every doctor knows a pandemic is coming, why is it not a single person penciled out the basics they would need?

Part of it is because the CMO has other responsibilities. Typically the responsibilities you note reside with the head of ID.

And we do know that. The issue is that everyone needs it at the same time. Normally, when our hospital is burning through PPE somebody else has it light (e.g. Australia). Normally epidemics don't cross all that many state lines and you can afford to count on "external resources" to cover the shortfall. But everyone got hit at once. China bought up a huge portion of the world's stockpiles while lying about the data.

As far as exact counts for PPE. Cannot be done. The amount of PPE a doc needs is going to be directly proportional to the amount aerosol generating particles the disease produces per patient per day. Knowing that would have required less lying from China. Likewise, the soiling rate is highly non-uniform against diseases, diseases that produce copious diarrhea need more gloves. Lung disease with heavy hemoptysis increase the number of face shields. Until you know the specifics of the disease (e.g. yes diarrhea, no projectile vomit, yes altered mental status) the amount and type of PPE cannot be known. The error bars are over an order of magnitude wide.

The closest you will come is for universal precautions, which we do have numbers on. Which we tried to ascertain if we needed. We tried to use the data to generate the exact numbers you want; we were at best misinformed.

And ultimately we come back to, again and again, that 95% of the deaths here should properly be laid at the feet of the CCP. They lied. Thousands have died. It is basically impossible to make accurate predictions in the face of such lying.

> The issue is that everyone needs it at the same time.

But everyone has been talking about worldwide pandemics forever. I just don't understand how those in charge (docs and administrators and professional orgs like AMA) didn't anticipate this. PS. I'd argue 100% of the deaths should be laid at the feet of the CCP. If we were experiencing this because of a screwup at a major company that had been experimenting with viruses and something nasty escaped from their lab we'd be putting people in jail.

The author of the article and the company he works for looks like a freight forwarder and import logistics manager.

His observation: "China is the only place in the world that can scale manufacturing as fast as we need right now. Our sources estimate the production capacity of Chinese PPE at 160M units per day. Of that, we believe 80M is consumed domestically and some is reserved for the Chinese national stockpile."

So, maybe the question to ask instead is: How much are you willing to pay for foreign sourced content during a pandemic when in the future there may be trade or other conflict.

Perhaps we should ask a different question: Should we signal that we will source domestically a certain percentage of demand and be willing to subsidize excess domestic capacity or pay for the quick conversion of domestic plants which currently produce raincoats to produce PPE.

Do price signals work when a foreign government might be concerned for its own persons first?

To me it seems the optimal is to allow backdoor price gouging and to avoid messy media leaks. Keep the process hidden like sausage making.

"Put another way, it’s the short run increase in price which allows long run elasticities to increase. Elasticity is endogenous to pricing."

Nobody cares for "elasticities" or "endogeneity" when they need a damn mask to go shopping for toilet paper. High minded nonsense is why Hillary failed to be elected.

"Credit constraints" are easily explained. Suppliers of medical equipment ought to be able to borrow at super cheep rates to meet orders or providers could borrow at super cheep rates to give down payments, if the Fed were targeting NGDP. But Fed has adopted a super tight monetary policy that has forced inflationary expectations below 1% pa over the next FIVE years. (Alex noted that, too --"Money is tight" -- in his piece above.)
Fed policy gave us an excruciatingly slow recovery from a purely demand side recession that did not require higher inflation expectations to keep on a constant NGDP growth path, just staying on the same old 2% PL path and it failed. When supply goes down, inflation expectations have to go UP to keep NGDP growing at the same rate. Could someone please explain this to Mr Powers?

I understand reputational reasons to explain why supermarkets do not dynamically price TP so as to always keep at least one roll on the shelf. The reluctance of suppliers in a B2B chain to nudge prices up in the absence of price controls is harder to understand. Could all the Libertarian grousing about price gouging be misplaced?

Back in the days of H1N1, countries all over the world placed massive orders for vaccines--hundreds of billions. Then it all kind of died out while the vaccines were being produced. And big countries like UK and France just walked away from their commitment to purchase. Yes, it went to courts. The countries argued it was a poor use of taxpayer money to buy something no longer needed.

If if I had one manufacturing line, it'd rather run that at max (3 shifts) and turn away business than scale and be left holding the bag when the music stops. And make no mistake: When the music stops, nobody will remember what they promised you. When people are experiencing pain, they will promise you anything.

The Flexport blog post is excellent. Unfortunately Alex seizes on the one element that is wrong. In this industry, capacity capex typically has a payback period of 1-3 years when demand exists. Assume that an order is paid upfront at 3x prices. That is nowhere near enough to enable a manufacturer to fund capex without external financing. Should prices rise? Yes. Would that solve capacity? Not at all.

Peterson is also wrong that medical distributors lack the cash flow to finance up-front purchases. While they are low-margin businesses with operating cash flows ~1-2% of revenues, they are extremely robust with easy access to credit lines and other forms of financing. Even under today's stressed conditions, they can find the cash. Hospitals, on the other hand, may have seen revenues drop too steeply to finance up-front purchases and may need A/P relief from their distributors. The subset of health systems that have internalized distribution may also struggle, but most of these have secondary contracts with distributors that they can utilize. That said, the actions of the federal government to intercept orders placed by states that flow through distributors may make hospitals hesitant to use this channel.

The real reason for delayed payment terms is that distributors and hospitals are low margin businesses with high incentives to generate cash flows from working capital optimization and the power to do so. Only very large medical manufacturers, such as Medtronic and BD, have the negotiating leverage to negotiate favorable payment terms. But, because manufacturers have higher margins and better cash flows than their customers, they have an incentive to provide value through payment terms before doing so through pricing.

The biggest missing element of this discussion is the role of group purchasing organizations (GPOs) in making prices inflexible. For commodity products like PPE, nearly all volume is sold at prices set by these GPOs on behalf of hospitals. The vendors are contractually prohibited from raising prices. Because these contracts typically go to the larger companies, it means that most supply is locked in to historical prices. Hospitals are allowed to purchase from other suppliers when they cannot get orders filled and those, typically smaller, vendors can sell at any price. Hospitals should have the incentive to pay those prices because the GPO contracts with vendors typically have clawback mechanisms where the contracted vendors are required to refund the difference between the contracted price and the price the hospital has to pay on the open market. I suspect that GPOs have been relaxing these terms, but haven't seen any reporting on this.

Finally, there is an interesting financing and investment dynamic in this product category. I've visited many Chinese, Mexican, and, rarely, US manufacturing plants for similar medical products. The Chinese firms, often operating on thinner margins, usually invest substantially more in capital equipment. They are more automated with bigger facilities and more spare capacity. I'm unsure whether this depends on easier financing supported by the Chinese state or simply a willingness to invest free cash flow to always be the lowest-cost producer rather than distribute it to shareholders. But, I suspect that the barrier to increasing capacity is more logistical than financial. The entire supply chain remains in chaos and increasing capacity requires buying equipment with long lead times. Those manufacturers need to build capacity, and the issue extends fractally all the way down the supply chain.

High quality comment. Thank you.

In my experience, the uncertainty with government decisions and the dangers of the mob in a blood-frenzy are much more important than the unwillingness of paying up-front. The channels are pretty much the same, so the retailers know the wholesalers, because they have been working for them for years. They know who to trust. I can give a personal process of decision.

In mid-March our pharmacy chain in Central America was offered 300.000 N95 masks. We had been in stock-out for a few weeks already. We estimated 300.000 masks would have lasted 10-12 days, if we rationed the sales to 5 masks per person. We had to pay cash, which is pretty usual with foreign suppliers, but we were certain there was no risk of compliance. Price was 3.5$, about 4 times the usual price (not that we had much experience with N95 masks, they are not usually sold in pharmacies, but we investigated the price chain). However, the price in the web was around 12$. We could have sold them at 4.5$ to cover costs, and create quite a lot of goodwill with our clients. However, at the end we decided not to go ahead, for two reasons. First, some newspaper or viral social media post would have attacked our brand, saying that we were selling at 3 or 4 times the usual price, greedily taking advantage of the “People”. Second, there was the risk that in the two-three weeks the masks needed to arrive to the markets and pass customs, the local governments would have frozen the price at the pre-pandemic level, or even appropriated the masks at the customs to give them to hospital workers and police.

Regarding the first concern, I cannot prove a negative of course, but it did happen to local producers of antiseptic gels. Regarding the second, in a couple of countries we operate the government froze the prices of surgical masks and antiseptics of any types. They did not froze the price of the N95s, but I am fairly sure it was just for bureaucratic ignorance, not intentionally.

Another example here in Italy, that I read in the press. Surgical masks have been in stock-out since February. A pharmacist in some ways found a few boxes of 50 units. She started to sell them two at the time, to give a chance to all her clients. One of them wanted an entire box and complained to the police, who gave the lady a 1000 euros fine, because she was not supposed to break-bulk the boxes.

Of course I am libertarian and very partial. But even the very statist among you, don’t you think that the States have been mostly a hindrance so far?

Very good point. I noticed this years back with Katrina, that the high prices weren't just "price gouging" they were what enabled the local retailers that remained open to pay the delivery drivers what the delivery drivers were demanding to drive into the hurricane zone. There's an entire supply chain that needs to be incentivized to deliver prduct where it's needed. The higher end-point price flows all the way back down the chain. It doesn't just go to the retailer at the end of it.

Another reason why higher prices are important even if supply is inelastic: we want the masks to go to the highest-valuing customers. If the price is high, that will happen.

Another reason why higher prices are important even if supply is inelastic: we want the masks to go to the highest-valuing customers. If the price is high, that will happen.

Two issues:

(1) "Value" may not mean what you think it means. The market measure of value is "willingness to pay" which is, of course, closely related to ability to pay. If you value the lives of lower middle class residents in Queens equally with rich upper west side folks in Manhattan, letting the Queens hospital be priced out of the market doesn't seem like "maximizing value." Similarly for the lives of Europeans versus the lives of Africans.

(2) We may still care about the distributional effects of prices. If we care more about consumer surplus versus producer surplus, the de facto transfer that occurs when price goes up may not be what we want. Insofar as purchases are paid for with taxes, you also want to think about the increased burden of taxes elsewhere in the economy to pay these higher taxes.

Yet another reason that higher prices are important is that they encourage people to seek out either substitute goods (KN95 vs N95 masks, for example) or ways to squeeze more use out of existing supply (sanitizing masks in an oven and reusing them up to 20x, for example).

Yes, but the blog makes excellent points about import controls as an obstacle to doing this. Add in regulations, and this isn't just a matter of price. The status quo situation is that a) nobody could legally import a KN95 mask because it was not approved by the FDA and/or not manufactured in an FDA-approved plant, b) hospital policies and procedures would prohibit this use and, c) medical litigation creates a very real financial risk for hospitals to be sued when a patient has a poor outcome and the hospital violated a or b above.

The point is that pricing is not the first-order problem here and is helpful, but nowhere near sufficient to remedy this shortfall. Fixating on price is just shoehorning the PPE supply dynamic into a prexisting viewpoint, not understanding the actual issue and finding the most important causes of the shortage. I would rank them:
1) global manufacturing capacity is not, and never would be, sufficient to handle this level of demand surge
2) increasing capacity requires financial incentives and coordination across the supply chain to accelerate capital investments
3) regulations prevent increasing capacity, substitution of like goods, etc.
4) contractual dynamics prohibit flexibility through the price mechanism
5) social opprobrium can hold back price increases, which can be helpful in sending signals regarding where supply should be directed

Maybe waiting to pay is still a good idea...


I'm sure availability of capital is an issue in some cases, but producers of N95 masks include companies such as 3M and Honeywell. Working capital is not the limitation on why they can't scale production more.

There's far too little discussion here of the inherent challenges to starting up new production lines even if capital is readily available. The somewhat quick solution to scaling production is to add more shifts if the line isn't already running 24/7. Even that can be a management challenge because of the need to have experienced people as managers and leads on additional shifts. I've known of companies that will keep a second shift operating at low volumes even when a recession reduces demand to levels that could be easily handled by one shift, in order to avoid the difficulties of having to set up a second shift again when volumes increase.

To that second point - even setting up a second line to double production capacity versus one line is a huge undertaking that typically takes months due to equipment lead time, hiring, and training. The idea that manufacturers could somehow scale up to 20x in a short period of time is pretty crazy.

GM and Ford making ventilators is the exception that proves the rule. These are huge organizations with large groups already in place for engineering, procurement, and production, as well as huge supplier bases. They were basically able to figure out how to take massive organizations and turn them into (by their standards) small ventilator manufacturers, but by their standards a "small" operation is very large compared to ventilator manufactures. It's an amazing accomplishment that they managed to do so in such a short period of time, but it's very different than a small organization growing to a much bigger one. I'll also note that we still don't know how the quality and reliability of the ventilators they produce will compare to ventilators from established manufacturers.


Nice share! thanks for the post in medical manufacturing.

Comments for this post are closed